Voice control

Post details

Touch screen phones are good for lots of things, but typing isn't one of them. On-screen keyboards take up too much room on the small screen, and are too small for touch typing.

Also, a small screen can only show a few options, so doing something quite simple may require three or four steps.

Voice control, or voice recognition, is one way to overcome these limitations. Here's an overview of the topic, followed by brief results from test with three apps for the iPhone and three for an Android phone.

How does voice control work?

There are three areas in which voice control can be used:

  • Dictating text: "dear john, thanks for your letter dated...."
  • Controlling apps: "send a text to john"
  • Searching the internet: "where is the nearest supermarket?"

Dictation is built into many phones, so apps are only required for the second and third of these areas.

Natural language processing is improving. While early voice recognition worked on each word in isolation, now whole phrases are recognised. This helps to know, for instance, when to interpret what you say as "there" or "their". And increasingly you can say "do I need an umbrella?" instead of "what is the weather forecast for today?".

Most apps require an internet connection because the intensive computing needed for voice recognition is not done on your device. Rather, what you say is sent to a more powerful computer online where the recognition is carried out, and the resulting text is sent back to your device.

On a computer, for best results you are advised to "train" a voice recognition programme to understand your voice. This requires you to read out quite long pieces of text which the computer analyses. Training is not needed for the simpler voice recognition used on a mobile device.

Different people may get different results from the same app, because it can be affected by the amount of surrounding noise, the way you talk and even the way you hold your device.

For best results, talk slowly and clearly directly into the microphone in a quiet area with a good internet connection. Using earphones for the output may also improve results by keeping spoken output separate from the voice commands you give.

New versions of computer and smartphone operating systems have some level of voice recognition built into them. It's also possible to get free or paid-for apps. We looked at a few of the apps available for iOS and Android phones.

Dictation

On the iPhone, iPod touch, iPad and newer Android devices, the on-screen keyboard has a key with a microphone symbol to the left of the Space key. Pressing this key starts dictation mode. You simply say what you want and then tap the screeen to indicate you've finished. Don't try reading War and Peace out in one go; dictation is intended for short pieces of text only.

Your device will try to work out what you've said and enter it into the active app.

It's important when dictating text that you vocalise capitalisation and punctuation, so you'd say "cap yes exclamation mark" to get "Yes!". The first word of a sentence and proper names may be capitalised automatically.

Android app tests

As with all things Android, not all apps work on all versions of Android, or on all handsets. The apps below ought to work on most recent handsets, but the Play Store will give you the definitive information.

1. Evi

Evi is free for Android devices, and is intended primarily for internet searches. It can also be used for sending emails, but not creating appointments or sending texts. Evi learns from your questions, and you can rate her answers.

If Evi doesn't catch what you say, or Google can't be reached, a message appears on screen which TalkBack doesn't read.

I wasn't impressed with the quality of Evi's answers - she could tell me my location and the weather, but not what the weather would be like tomorrow in a different location, of which supermarket was nearest to me.

2. Jeannie

Jeannie is also free, and has a far greater range than Evi - she can send emails and texts, make calls and set alarms, as well as conduct internet searches.

I found Jeannie quite slow, but her answers are almost all short pieces of text which she reads out perfectly. She understood what I said and gave correct answers most of the time, although she thinks Tony Blair is still the Prime Minister! She has an accessibility mode to "optimize workflow for VI" but doesn't make it clear what this actually does, and there is a useful "Shake to Wake" option.

3. Voice Actions

Voice Search is Google's voice recognition for Android. It's free and it allows you to send texts or emails, call contacts, listen to music, get directions, go to websites or write notes. Oh, and you can also search Google.

The recognition is built into the Google Search widget that appears by default on many Android devices. The widget is a bar with the word Google at the left and a microphone image at the right. Tapping the microphone image starts the voice recognition.

It's very fast, but was quite picky about how commands are phrased - for example, "create an appointment tomorrow at 3pm" didn't work, while "create an appointment at 3pm tomorrow" did - and it refused to enter any punctuation in an email!

iOS app tests

The good thing with iOS is that providing the user selects any available update all devices are updated to the same version, regardless of whether it is an iPhone, iPad or iPod. This means that the apps will work across all devices.

1. Evi

Evi is free for iOS devices and, like its Android relative, is intended primarily for internet searches. It can be used for sending emails and also text messages. Again Evi cannot create appointments.

I quite liked the answers Evi gave. It can take some time for it to recognise what you are saying, which can be frustrating, but the clarity was impressive when she did get it right. The description of the location was very good telling me what my nearest landmark was and offering a map.

When used with VoiceOver the Evi voice is disabled.

2. Vlingo

Vlingo is free for iOS devices. The first time you use it a short tutorial comes up which tells you how to use it and what voice commands to use. I found it unnecessary in most cases as it generally understood what I was asking of it.

Vlingo will allow you to send email and text messages, use maps, search the internet, update twitter and Facebook as well as make telephone calls.

Vlingo generally understood what it was that I was asking it however it became completely unusable with VoiceOver. It would recognise "send an email" but just open a blank email which I had to fill in using gestures.

3. Siri

Siri is pre installed on all iOS devices running version 4 and above. Like Google's Voice Actions, it's free and allows you to send texts or emails, call contacts, listen to music, get directions, search the internet and see what the weather is like anywhere in the world.

Siri is very simple to use and occasionally there is an issue if you are not clear enough due to a cold but on the whole Siri is very user friendly and accurate. The massive benefit of Siri is that it's activated by holding down the home button and will work even if the device is locked. I love this most in the winter when I want to make a call and don't want to take my gloves off!

Conclusion

Hopefully the information above gives you a flavour for where voice control of devices is at the moment. Undoubtedly a year from now, the situation will have moved on a long way.

As well as the above apps, there are some other Android apps it will be interesting to track:

  • Edwin:specifically designed to provide speech output as well as speech input, but it hasn't been updated in two years, which doesn't bode well.
  • utter! Voice Commands: very impressive, out as a beta (i.e. not yet finished) and the developer is working on its accessibility.
  • Dragon mobile assistant: out in the US in January 2013, but not available in the UK yet. From the makers of the "daddy" of voice recognition programmes for computers.

There doesn't seem to be the same number of iOS apps out there, maybe because the built-in recognition makes them redundant!

If you use any other voice control apps, we'd love to hear about them.



Comments (0)

Tag: Technology news

Posted at: 21/02/2013 11:00 AM by Steve Griffiths

There are no items in this list.

Add a comment

If you would like to add a comment, you need to login first.